Blog: Pairing the Best Minds with Advanced AI to Protect the Planet
With floods, droughts and extreme weather events, the ravages of climate change seem to be in the news daily. While the best scientific minds on the planet are trying to address these and other environmental problems, data scientists have enlisted AI technology to help tackle them as well.
AI is ideally suited to support the work of scientists in addressing these issues because it is based on the same methods that scientists use. Data scientists similarly start with a problem or question that they want to answer, and build an AI program around a hypothesis that they are trying to test — the difference is that with AI they can now crunch huge amounts of data to shed insight into the problem. Since AI can process this data exponentially faster than humans, these programs are identifying patterns that can be used to get to the bottom of issues so much faster than before and even predict future outcomes based on historical data.
Thanks to powerful supercomputers that are processing massive amounts of data at a faster pace than ever before, IoT technology is enabling data scientists to collect real-time, continuous data from the field to paint a more complete picture of the environment, and gather vital data to train the AI programs. For example, by using IoT sensors to collect data on plant moisture, soil samples and temperature on an ongoing basis, farmers can optimize the best times for planting, spraying, watering and harvesting their crops to reduce pesticide, fertilizer and water usage, among other benefits.
Imagery from satellites, drones and the field is providing additional data that AI programs can use to gain insight into environmental issues. For example, to help prevent mosquito-borne illnesses in Puerto Rico and elsewhere, the Puerto Rico Vector Control Unit (PRVCU) is classifying thousands of mosquitos, across Puerto Rico for research and testing. An AI-based program is automating the mosquito identification and classification process based on the training data captured in the lab and labeled by a private crowd. Since the program automates and significantly speeds up this process, it frees scientists from doing it manually, saving months of effort, and enabling scientists to instead focus on identifying the causes of insecticide resistance and helping to prevent the spread of these diseases.
Here are seven other ways AI is supporting scientists in helping the environment:
• Better weather forecasting. Climate Informatics is an emerging area of study, enabling forecasters to more accurately predict weather, including severe weather events. That way, authorities, for example, can more accurately issue early warning alerts to the general population about these impending events, increase preparedness and boost disaster response.
• Cleaning up plastic in the ocean. AI is helping an autonomous floating system sweep up plastic garbage in the ocean. The Ocean Cleanup organization estimates that by using about 60 of these systems about half of the debris in the Great Pacific Garbage Patch can be removed in 5 years.
• Monitoring vegetation following a hurricane. Professors from Columbia University and the Data Science Institute are using AI to study the impact Hurricane Maria had on Puerto Rico’s El Yunque National Forest. Using tree identification data and aerial images, the AI program is able to determine the damage caused by the hurricane and changes that occurred in the tree species.
• Fighting air pollution. Sensors and AI software can monitor air quality in real time. Also, by identifying activity such as crop burning, which is common in areas like Delhi, India, the AI programs can alert authorities to problems that need to be addressed. In addition, using AI-developed models, scientists can project the outcomes of different preventative measures to fight air pollution.
• Creating smarter routes. By finding the best and fastest trucking routes that reduce idling at lights, AI programs can help reduce carbon emissions.
• Monitoring wildlife. Images of giraffes collected from their habitat are analyzed by AI software to identify individual animals by their patterns. This helps wildlife experts estimate the population and giraffe movement to help protect the species.
• Keeping tabs on the Earth. Using NASA-collected images of the Earth’s surface, AI programs are able to “see” the changes that are occurring and develop more accurate climate change models.
We’re only at the tip of the iceberg when it comes to putting AI to work for the good of the planet, but the possibilities are endless. Thanks to many academic, private and public initiatives, such as Microsoft’s AI for Earth, which provides grants for initiatives that use AI technology to heal the Earth; or AI for Good, a worldwide public charity, scientists and data scientists are working together to create positive environmental change. Through these many initiatives the scientific community is seeing the value of applying AI to solve nature’s challenges and how it just might be the spark needed to help ensure clean water, the continued abundance of crops and the proliferation of wildlife.